Affiliations: Cornell University, Sibley School of Mechanical and Aerospace Engineering, Ithaca, NY, USA
Corresponding author: Jacopo Banfi, Cornell University, Sibley School of Mechanical and Aerospace Engineering, 541 Upson Hall, Ithaca, NY 14853, USA. E-mail: [email protected].
Note:  This article is based on the PhD thesis of the author, carried out at the Politecnico di Milano, for which he was awarded the “Marco Cadoli” award from the Italian Association for Artificial Intelligence.
Abstract: In the multirobot exploration problem, a team of robots is deployed in an unknown environment with the aim of efficiently mapping free space and obstacles. The study of this problem is often simplified by assuming that communication between robots is possible between any two locations. In practice, this could be a far too optimistic assumption due to the limited range of the robots’ local communication modules. This paper presents, under a unified framework, a series of recent contributions in the field of multirobot exploration under communication limitations addressing three of its key research challenges. First, the development of an exploration strategy able to enforce timely reports at a supervising control station without excessively hindering the exploration process. Second, the construction of “communication maps” of the environment: maps that –when queried by the robots to determine the possibility of a future communication with a teammate–return the estimate of the radio signal strength between two given locations, along with a numerical value representing the confidence in the prediction. Third, the problem of computing backup reconnection plans to be used in presence of mispredicted communication links. The paper also discusses some promising research directions derived from the lessons learned.
Keywords: Multirobot systems, multirobot exploration, communication maps, multirobot reconnection, planning and scheduling